Designing less noble yet high-performing multimetallic catalysts for catalytic converters using a high-throughput approach

Abstract

The development of efficient yet cost effective multimetallic three-way catalysts for automotive emission control presents significant challenges, requiring a consistent dataset of adequate size to expand the compositional search space and build up a systematic understanding of complex multielement interactions. The present study implemented a high-throughput approach for consistently synthesising and evaluating 140 multimetallic three-way catalysts. The results highlight the importance of synergistic interactions among multiple metals, leading to markedly improved catalytic performance compared to monometallic platinum group metal (PGM) catalysts. The results revealed that PGM ratio can be an efficiently tunable parameter, leading to more cost-effective formulation with comparable or even superior activity in specific combinations, such as Ni–Pd–Ag–Pt–Pt, Cu–Pd–Pd–Pd–Ag, and Cr–Ru–Pd–Pd–Ag, with optimal NO reduction performance (T₅₀ = 271–296 °C) at 40–60% PGM contents. In addition, the comprehensive dataset enabled machine-learning-aided analyses and statistical validation for understanding the catalyst system. SHAP analysis confirmed promoting elements such as Pd, Rh and Ni, with their synergistic combinations providing statistically significant enhancement (p < 0.05). These findings underscore the power of combining high-throughput experimentation with data-driven approaches for designing next-generation three-way catalysts that balance performance, cost and sustainability for automotive emission control.